These 3D printer test print files are made in your browser from the sizes you choose, so they fit your printer and test exactly what you want. There are eight original designs: a calibration cube with axis notches, an overhang test from 20 to 70 degrees, a bridging test, a tolerance test with pins printed in their holes, a stringing test with two towers, a first layer test sized to your bed, a temperature tower with each block's temperature engraved on it, and a ringing tower. Every file is a closed, watertight binary STL that loads cleanly in Cura, PrusaSlicer, OrcaSlicer, Bambu Studio and Creality Print.
How to use the 3D printer test prints
- Pick a test print on the left. The 3D preview shows it on a 10 mm grid, with the orange line along X and the green line along Y.
- Set the sizes. For the first layer test, type your bed size. For the temperature tower, type the temperature range for your filament.
- Press Download STL, or All 8 prints to get every file in one ZIP.
- Slice it with your usual profile and change only the slicer settings shown under the preview.
- Print it, then compare the result with the matching section below.
Change one setting at a time and write it on the print with a marker. When two settings change at once, you cannot tell which one fixed the problem.
What each test print shows
| Test print | What it checks | A good result |
|---|---|---|
| Calibration cube | Size on X, Y and Z, elephant foot | Within 0.1 to 0.2 mm of the set size |
| Overhang test | The steepest unsupported slope | Clean undersides up to 45 to 60 degrees |
| Bridging test | Spans across empty space | Tight, straight strands on 20 mm or more |
| Tolerance test | Smallest gap that stays free | Pins at 0.2 to 0.3 mm push out cleanly |
| Stringing test | Retraction | No hairs between the towers |
| First layer test | Bed level and Z offset | Five smooth, flat squares |
| Temperature tower | Best nozzle temperature | Glossy blocks with clean bridges and no strings |
| Ringing test | Vibration after corners | Flat walls after each notch |
Strings and hairs between parts
Fine hairs between the stringing towers mean molten plastic oozes from the nozzle while it travels. In order of likelihood:
- Retraction too short or too slow. Most direct drive printers need 0.5 to 2 mm at 25 to 45 mm/s, and most Bowden printers 4 to 7 mm at 40 to 60 mm/s. Raise the distance in 0.5 mm steps and print the towers again. The 3d printer calibration page has a retraction tower planner.
- Nozzle too hot. Lower it 5 °C at a time, or print the temperature tower to find the coolest clean block.
- Damp filament. PETG, nylon and TPU absorb moisture from the air and string badly when wet, often with popping sounds. Dry the spool at the temperature its maker lists, usually 4 to 6 hours.
- Long travel moves. Turn on combing in Cura or avoid crossing perimeters in PrusaSlicer for real prints.
Gaps, thin walls and weak parts
Under extrusion shows as gaps between lines, thin or missing walls and a cube that snaps easily. Check that the spool turns freely and is not tangled, then listen to the extruder: clicking means the gear is slipping because the nozzle is partly blocked or the speed is too high for the temperature. Clean or replace the nozzle, raise the temperature 5 °C or slow down. If the printer is mechanically fine, calibrate the extruder steps (E-steps or rotation distance) and the flow rate in that order.
The bottom layers bulge out (elephant foot)
A lip around the first few layers of the cube is elephant foot. The first layer is squashed too hard or stays soft on a hot bed. Raise the Z offset by 0.02 to 0.05 mm, lower the bed temperature by 5 °C, and use the slicer's compensation: Elephant foot compensation in PrusaSlicer and OrcaSlicer, or a small negative Initial Layer Horizontal Expansion in Cura. Measure the cube above the lip for its true size.
Ripples after corners (ringing)
Ringing, also called ghosting, is a fading echo of each corner and notch on the walls that follow. The frame or print head vibrates when it changes direction fast. Tighten the belts and any loose frame bolts, and check the wheels on the print head for wobble. Then lower acceleration, or use input shaping: Klipper has it through its resonance tools, Marlin 2.1.2 and later has it with the M593 command, Bambu Lab printers tune it themselves, and recent Prusa printers ship with it calibrated at the factory. The ringing tower's frequency calculator turns your ripple count into the number those settings need.
Overhangs droop or curl up
Note the steepest fin on the overhang test that printed cleanly. Drooping or curled edges past that angle come from too little cooling, too much heat or too much speed. Run the part fan at 100% for PLA, slow down the outer walls, lower the temperature and try a thinner layer height. Use supports beyond the angle you measured. The fins are labelled from vertical, like Cura's Support Overhang Angle. PrusaSlicer, OrcaSlicer and Bambu Studio set the support threshold from horizontal, so use 90 minus your result there.
Corners lift off the bed (warping)
Warped corners on the cube or the tolerance plate come from the plastic shrinking as it cools. Wash the bed with dish soap and warm water, or wipe it with isopropyl (IPA), since finger grease ruins adhesion. Check the first layer test, raise the bed temperature (typical values: PLA 55 to 65 °C, PETG 70 to 85 °C, ABS and ASA 95 to 110 °C), add a brim, and keep drafts away. ABS and ASA need an enclosure.
Layers shifted sideways
A sudden step in the layers means the print head lost its position. The usual causes are the nozzle hitting a curled up part, a loose belt or pulley grub screw, or speed and acceleration higher than the motors can deliver. Tighten the belt so it twangs like a low guitar string, check the pulley screws, and lower the speed. If shifts happen late in long prints, the motor drivers may be overheating, so check the cooling fan on the electronics.
Using the test prints on a resin printer
Resin printers need different settings, but some of these prints still help. The tolerance test shows exposure: overexposure makes holes smaller and pins fatter, so small gaps fuse. The calibration cube shows whether the printer prints true to size. Print the overhang and bridging tests only with supports, since a resin printer cannot print into empty space. Most resin slicers and many printers also include an exposure test that prints the same pattern at several exposure times, which is the right place to start. For layers that separate, lower the lift speed and raise the normal exposure. For prints that fall off the plate, raise the bottom exposure and keep the resin above 20 °C.
Looking for a 3D printer test page?
There is no single test page for a 3D printer like the one a paper printer makes. The first layer test and the calibration cube together are the closest thing. If you want to check a paper printer instead, print a test page. To send commands such as M503 straight to a printer connected by USB, use the serial monitor online, and if you watch your prints through a USB camera, check it with the webcam test.
What the generator cannot do
The STL file holds only the shape. Speed, temperature and retraction come from your slicer, so the same file can print well or badly. The preview shows the model, not how it will print, and the generator cannot see your printer. Use calipers for the measurements: printed parts usually vary by 0.05 to 0.1 mm, so judge trends over several prints rather than one reading.
3D printer test print questions
What is the best first test print for a new 3D printer?
Start with the first layer test, because almost every other fault begins with a bad first layer. Then print the calibration cube to check size and the stringing test to check retraction. A temperature tower is worth printing for every new brand of filament.
Is the 3D Benchy a good test print?
Benchy is a popular all round torture test, but when it fails it does not tell you which setting is wrong. The prints on this page each check one thing, so a failure points straight at the setting to change.
How long does a test print take?
Most take 10 to 45 minutes at normal speed: the cube about 20 to 40 minutes, the first layer test under 15. The temperature tower takes the longest, often 1 to 2 hours for seven blocks.
Can I download G-code instead of STL?
No. G-code is made for one printer, nozzle and filament, so a ready made file would print wrongly on most machines. Open the STL in your slicer with your usual printer profile, then change only the settings listed in the tool.
Are these test prints free to share?
Yes. They are original designs made in your browser from your settings, so you can print, share and modify them freely.
What does a calibration cube tell you?
It shows whether your printer prints true to size on each axis, and it shows elephant foot at the bottom and ringing on the walls. Measure it with calipers. A cube off by more than about 0.2 mm on a 20 mm side is worth a look.